The Automation Core

EV Winter Range Loss Explained: Heat Pump vs PTC Heater Test Data and the Cold-Start Catch

· battery thermal management, cold weather EV, EV preconditioning, EV winter range, heat pump, PTC heater, regenerative braking, The Engineering Core,

 


Bottom line: After my breakdown of how much A/C costs an EV in summer, the obvious follow-up is winter, and winter is worse. Heat pumps help, but not where you might expect. Three data sets explain why: Norway's El Prix winter test, a same-model heat pump vs PTC dynamometer study, and a regen-to-heat study reported on October 8.

What happened

  • El Prix Winter 2026 (NAF and Motor, Norway, January): −7 to −32°C, the coldest edition yet. Cars missed their WLTP range by 38% on average, versus 18% in 2025 at −6 to +8°C. The winner, a Lucid Air, covered about 520 km against a 960 km WLTP rating, a 46% gap.
  • Heat pump vs PTC on a dyno (World Electric Vehicle Journal 2026, 17(4)): two cars of the same make and model, a 2022 model year with a heat pump and a 2020 model year with a PTC heater, run over standard cycles at −10, −7, 0 and 25°C. Averaged over the cold temperatures, the heat-pump car retained 7% more range on UDDS, 7% on HWFET and 4% on US06.
  • The advantage peaked at 0°C (15% on UDDS) and shrank to 1–5% at −7 and −10°C. At −10°C, on a cold start the heat pump used more energy than the PTC heater; it only pulled ahead once the cabin was warm.
  • Reported October 8: a study in the journal Ionics used logs from 72 real BMW i3 trips to explore routing regenerative-braking energy to cabin heating, plus machine-learning models that predict end-of-trip state of charge.

Back-of-envelope (my illustrative numbers, not study data): say the cabin needs 3 kW of heat. A PTC heater is roughly 1:1, so it draws 3 kW. A heat pump at a COP of 2.5 draws 1.2 kW. In 30 km/h city traffic at 0.13 kWh/km, traction needs only about 3.9 kW, so PTC heat adds 77% and the heat pump 31%. At 100 km/h and 0.18 kWh/km, traction is about 18 kW and the same 3 kW adds just 17%.

1. Why winter hurts more than summer

HVAC load scales roughly with the indoor-outdoor temperature difference. Cooling from 35°C to 24°C is an 11°C delta; heating from 0°C to 22°C is 22°C, and from −10°C it's 32°C. Stack on a cold pack with higher internal resistance and less usable energy, plus defogging, and the losses compound. A combustion engine heats the cabin with waste heat it throws off anyway. An efficient e-drive leaves very little to borrow.

2. The heat pump's cold-start catch

A heat pump moves heat from outside air into the cabin. The colder the air, the less heat there is to move, so COP falls, and many systems lean on a PTC backup. On a cold start, the refrigerant loop and cabin are both frozen, and the system spends energy warming itself up first, which matches the dyno result at −10°C. The practical takeaway: heat pumps pay off most around 0°C and least on very cold, very short commutes, which is exactly where drivers feel winter range loss the most.

3. A cold battery can't absorb much regen

Charging lithium-ion cells in the cold risks lithium plating on the anode, so the BMS caps charge current, and regen gets throttled. The Ionics study's idea is to send recovered energy straight to the heater instead of forcing it into a cold pack, skipping a charge-discharge round trip and a battery cycle. The coverage says both range and battery life improved, but gives no magnitude. The headline R² of 0.84 measures how well the model predicts remaining charge, not extra kilometers.

4. Winter settings that actually work

  • Precondition while plugged in, so grid power, not the pack, warms the cabin and battery.
  • Heat people, not air: seat and steering-wheel heaters draw tens to a couple of hundred watts, versus kilowatts for the whole cabin.
  • Precondition before DC fast charging: a cold pack charges slowly, and navigating to a charger usually triggers battery warm-up.
  • Check tire pressure: it drops with temperature, and rolling resistance rises with it.

Open questions

  • The dyno study compares two model years of one vehicle; other differences between them may affect the result.
  • The Ionics study doesn't publish a range gain, and its data covers 72 trips in one small car.
  • El Prix is a one-off road run whose route, speed and stop pattern differ from a daily commute.
  • Heat pump designs vary widely (waste-heat recovery, multi-way valves, vapor injection), so results don't transfer neatly between brands.

Related videos (The Engineering Core)

Your turn: how much range does your EV lose in winter, and do you rely on preconditioning or just seat heaters? Share your model and numbers in the comments.

← Newer articles Older articles →